We're looking for a senior engineer to own the JVM platform that the rest of engineering builds on, and to point it at one goal: making it easy for internal teams to ship LLM-powered agents.
Concretely, that means implementing in the Kotlin and Spring Boot ecosystem: reusable libraries, Spring Boot starters, templates, and shared building blocks that teams depend on. The difference from a classic platform role is what those foundations are for. Agents need message routing, tool exposure, memory and state handling, human-in-the-loop workflows, MCP integration, and evaluation and guardrail hooks. The team, which is internally called the BASE team, focuses on turning it into a platform.
This is a high-leverage, builder-and-multiplier role. You'll make foundational design decisions, set the patterns and standards other engineers follow, and raise the bar for everyone building on top. Because your foundations sit underneath many teams' work, understanding how the company fits together (its products, domains, and the teams behind them) is central to doing the job well.
You don't need to arrive as an agent expert, but you have to be interested in understanding the concepts, learn and provide reusable solutions in that field in the long run.
What You'll DoOwn the design of shared libraries, Spring Boot starters, and templates, making the architecture, API, and trade-off decisions that determine how easily internal teams ship on top of them.
Build the JVM-side foundations agents run on: message routing, tool and capability exposure, state and memory handling, human-in-the-loop workflows, MCP integration, and the surrounding plumbing.
Drive technical work across our Kotlin/Spring Boot services and the client libraries that wrap them.
Take part in defining the evaluation and guardrail strategy so we can measure agent quality and reliability and catch regressions before they reach customers.
Set the patterns and standards for how teams build agentic systems here, and bring those teams along with them.
Build and maintain a clear map of the wider ecosystem, meaning how our systems, domains, and teams connect, and use it to make foundational decisions that fit real needs across the company.
Drive adoption and multiply other engineers: Mentor teams, review designs and code, own the docs and examples that turn a good framework into one that teams actually choose, and feed what you learn back into the platform.
Raise the bar on developer experience: clear APIs, good defaults, useful errors, documentation people actually read, and hold the platform to it.
5+ years of professional software engineering experience, including a track record of building libraries, frameworks, or platform components that other engineers built on.
Deep experience with Kotlin and/or Java on the JVM, and with Spring Boot, plus strong instincts for API and library design. You know what makes a foundation pleasant to build on versus painful.
Demonstrated technical leadership: you set direction, make and justify architectural trade-offs, and lift the engineers around you, with or without formal authority.
Genuine drive to lead on agentic AI and its surrounding tooling: orchestration, evaluation, guardrails.
The judgment to drive ambiguity to clarity: to take an open-ended or unfamiliar problem and turn it into a well-scoped design others can execute.
A systems thinker who owns the bigger picture: you actively understand how the company's products, domains, and teams connect, and design for the whole, not just the module in front of you.
Good cross-team communication: you are able to work and communicate with multiple teams as we are providing the baseline of their development.
Working familiarity with AWS (or any Cloud) and Kubernetes: enough to reason about how your libraries behave once deployed.
These aren't entry requirements, but they're what will make you effective fastest:
Hands-on experience building agents and agent graphs: multi-step orchestration, tool use, memory, and the failure modes that come with non-deterministic systems.
Practical Python skills: our agent-side tooling is Python, so you should be comfortable enough in the language to prototype, read others' code, and act as a senior voice across both stacks. Deep experience running large-scale Python in production isn't expected.
Experience defining evaluation, experiment-tracking, or observability strategy for ML or agent systems.
Distributed systems background: messaging and queues (SQS, Redis), gRPC, event-driven architectures, including the failure modes and operational realities.
Deeper AWS and infrastructure-as-code experience.
Experience growing a platform or DevEx practice: standards, golden paths, and adoption across multiple teams.
Contributions to or maintenance of open-source libraries or frameworks.
